Generation of Human Cardiac Organoids from Embryonic Stem Cells via Stepwise Mesoderm Induction and 3D Self-organization

Cardiovascular disease remains the leading cause of morbidity and mortality worldwide, underscoring the urgent need for reliable in vitro models that recapitulate human cardiac development and function^1,2. Traditional two-dimensional (2D) monolayer cultures of cardiomyocytes, while valuable, fail to reproduce the spatial complexity, multicellular interactions, and electromechanical integration of the native human heart^3,4. 3D cardiac organoids derived from pluripotent stem cells have emerged as powerful platforms that address these limitations, providing self-organizing, multicellular structures that mimic early cardiac morphogenesis and functional maturation in vitro⁵.

hESCs, such as the H9 line, possess the intrinsic ability to differentiate into all somatic cell types, including cardiomyocytes, endothelial cells, and cardiac fibroblasts⁶. Under controlled culture conditions, hESCs can be guided through the sequential stages of mesoderm induction, cardiac mesoderm specification, and cardiomyocyte differentiation^5,7. This stepwise process can be regulated using specific signaling molecules and inhibitors, including Activin A, BMP4, FGF2, PI3K inhibitors, CHIR99021, Wnt inhibitors, and retinoic acid^5,8. By closely mimicking the in vivo signaling environment of embryonic cardiac development, these cues promote efficient and reproducible cardiogenesis in vitro^5,7.

Unlike many organoid systems that require embedding in extracellular matrix (ECM) components such as Matrigel, cardiac organoids can self-assemble in low-adhesion conditions without matrix support^5,9. Aggregation of differentiating hESCs in suspension yields spherical 3D structures that exhibit spontaneous beating within the first week of differentiation¹⁰. These organoids contain functionally relevant cardiac cell populations and demonstrate characteristic features, such as sarcomeric organization, calcium transients, and rhythmic contractions^11,12, offering a physiologically relevant model of early human heart tissue.

The ability to generate hESC-derived cardiac organoids offers exciting opportunities for developmental biology, disease modeling, drug screening, and cardiotoxicity testing^13,14. Moreover, integration with genome editing technologies and patient-specific induced pluripotent stem cells (iPSCs) enables personalized disease modeling and therapeutic exploration for inherited cardiac disorders.

In this article, we present a detailed, step-by-step protocol for the generation of cardiac organoids from human embryonic stem cells, highlighting key stages of differentiation, culture conditions, and validation techniques. This platform provides a robust, scalable system for recapitulating human cardiac development and establishing a foundation for future translational research.

All procedures involving human embryonic stem cells (H9) were conducted in accordance with institutional guidelines and approved by Biomedical Ethics Committee of Southwest Medical University, approval 20241023-031. All tissue culture work detailed below should be done in a Class II laminar flow hood. Always ensure that media and reagents are at room temperature, equilibrated naturally before use. Do not use a water bath to warm media.

1. Thawing and routine culture of HESCs (H9)

1. Vitronectin coating of culture plates
   1. Thaw Vitronectin stock on ice and dilute to a working concentration of 10 µg/mL using cold E8 medium.
   2. Add 1 mL per well of the diluted Vitronectin solution to a 6-well plate. Incubate the plate in a 37 °C, 5% CO₂ incubator for 1-4 h before use.
   3. Prior to seeding cells, aspirate excess Vitronectin and do not rinse.
      NOTE: Incubation in the cell culture incubator improves coating uniformity and enhances cell attachment, especially important for sensitive hESCs. Do not allow the coating solution to dry. This protocol uses vitronectin-coated plates only during the hESC thawing and expansion stages to provide a defined, surface-bound ECM substitute. Cardiac organoids are subsequently formed through self-aggregation in low-adhesion suspension conditions, without the need for ECM.
2. Thawing cryopreserved HESCs (H9)
   1. At least 30-60 min in advance, allow all required media and reagents (e.g., E8 medium, E8 + ROCK inhibitor, PBS) to equilibrate naturally to room temperature.
   2. Retrieve one cryovial of H9 human embryonic stem cells from liquid nitrogen storage using appropriate personal protective equipment. Immediately immerse the cryovial in a 37 °C water bath. Gently swirl the vial and monitor closely.
   3. Once a small ice crystal remains (~90% thawed, ~1-2 min), promptly transfer the vial into a biosafety cabinet. Disinfect the outside of the cryovial with 75% ethanol before opening.
   4. Slowly transfer the contents of the cryovial dropwise into a 15 mL conical tube containing 5 mL of room temperature E8 medium supplemented with 10 µM Y-27632 (ROCK inhibitor). Mix gently by pipetting up and down with a wide-bore pipette tip.
   5. Centrifuge at 200 × g for 5 min at room temperature. Carefully aspirate the supernatant and resuspend the cell pellet in 2 mL of E8 + Y-27632 medium.
   6. Plate cells onto a Vitronectin-coated 6-well plate (prepared in advance) at the desired seeding density (typically 150,000-200,000 cells/well). Gently rock the plate to distribute the cells evenly, and transfer to a 37 °C, 5% CO₂ incubator.
      NOTE: The use of ROCK inhibitor (Y-27632) is critical during the first 24 h post thaw to prevent anoikis and support survival.
3. Routine maintenance and passaging
   1. Change medium daily using 2 mL of fresh E8 medium per well.
   2. Monitor daily under a phase-contrast microscope; passage cells when they reach ~70-80% confluency, typically every 3-4 days.
      1. To passage: Aspirate medium and rinse gently with 1 mL of DPBS.
      2. Add 0.5 mL of stem cell dissociation reagent and incubate at room temperature for 3-5 min. Tap the plate gently to dislodge colonies.
      3. Collect the cell suspension and replate at a 1:3-1:4 split ratio onto freshly Vitronectin-coated wells in E8 medium containing 10 µM Y-27632. After 24 h, replace the medium with E8 without ROCKi.
         NOTE: For optimal recovery and subsequent differentiation efficiency, HESCs should be passaged as small clusters of 2-10 cells.
         Use of wide-bore tips is recommended to minimize shear stress. A 5 mL pipette tip or a 10 mL serological pipette can be used for mixing.

2. Differentiation of HESCs (H9) into cardiac lineage

1. Induction of mesodermal differentiation (Day 0-1.5)
   NOTE: Avoid directly differentiating freshly thawed HESCs (H9) into cardiac organoids. It is crucial that the cells are first passaged and maintained for at least one passage before initiating differentiation to ensure optimal cell viability and healthy differentiation. Differentiating cells immediately post thawing can result in poor outcomes due to compromised cell health.
   1. Preinduction seeding: When hPSCs reach approximately 70% confluency, gently dissociate the cells into small clusters using an appropriate dissociation reagent. Resuspend the clusters in 2 mL of E8 medium supplemented with 10 µM ROCK inhibitor (Y-27632).
   2. Aggregation: Seed 200 µL of the cell suspension (typically 12,000-15,000 cells) into each well of a 96-well ultra-low attachment plate and then centrifuge the plate at 200 × g for 3 min at room temperature to facilitate cell aggregation at the bottom of each well.
   3. Incubation: Place the plate in a 37 °C, 5% CO₂ incubator and culture for 24 h to allow formation of compact and uniform embryoid bodies (EBs).
   4. Prepare Basal Culture Medium (BC-M) by combining 50% IMDM, 50% F12 nutrient mixture, 15 µg/mL of Insulin-Transferrin-Selenium (ITS-G), 450 µM of monothioglycerol, and 5 mg/mL bovine serum albumin fraction V.
   5. Prepare MD-M by supplementing BC-M with Activin A (50 ng/mL), BMP4 (10 ng/mL), CHIR99021 (3 µM), FGF2 (30 ng/mL), and LY294002 (5 µM).
   6. Mesodermal Induction: After 24 h of aggregation, induce EBs for 36 h with MD-M.
      NOTE: To ensure the induction efficiency and consistency of cardiac organoids, we strictly limited the use of H9 hESCs to within three passages after thawing. This strategy is critical for maintaining high differentiation efficiency and overall experimental reliability.
      Gentle handling during medium exchange is critical to preserve the integrity of the formed aggregates. Avoid disrupting the spherical morphology of the EBs.
2. Induction of cardiac mesodermal differentiation (Days 1.5-5.5)
   NOTE: The Wnt signaling inhibitor (XAV-939) is critical to suppress posterior mesodermal signals and promote cardiac lineage specification. Retinoic acid plays a dose-sensitive role in anteroposterior patterning; excessive RA may lead to foregut specification.
   1. Supplement BC-M (composition detailed in Step 2.1) with BMP4 (10 ng/mL), FGF2 (10 ng/mL), Retinoic Acid (0.5 µM), and XAV-939 (5 µM) to prepare Cardiac Mesodermal Differentiation Medium (CMD-M).
   2. After 36-40 h of mesodermal induction, carefully aspirate the MD-M and replace with CMD-M for 4 days with medium change every day.
      ​NOTE: Handle aggregates gently during medium changes to avoid mechanical disruption or stress-induced apoptosis.
      ​During this period, spheroids become more compact and organized. Early signs of cardiac lineage commitment can be observed under brightfield microscopy by Day 5.
3. Cardiac organoid specification (Days 5.5-7.5)
   1. Supplement BC-M (composition detailed in Step 2.1) with BMP4 (10 ng/mL), FGF2 (10 ng/mL), and insulin (10 mg/mL) to prepare cardiac organoid specification medium (COS-M).
   2. On Day 5.5, gently aspirate CMD-M and replace with COS-M for 2 days with medium change every day.
      NOTE: BMP4 and FGF2 synergistically promote cardiac progenitor expansion and suppress non-cardiac fates at this stage. Insulin supports metabolic activity and structural development of differentiating cardiomyocytes.
4. Cardiac organoid maturation
   NOTE: Avoid full dissociation or pipetting of spheroids during transfer. Cardiomyocytes are highly sensitive to shear stress.
   1. Supplement BC-M (composition detailed in Step 2.1) with insulin (10 mg/mL) to prepare Cardiac Organoid Maturation Medium (COM-M).
   2. From Day 7.5 onwards, transition to COM-M. Perform medium changes daily to maintain nutrient levels and remove metabolic waste while minimizing stress.
   3. From Day 8 to Day 12, look for spontaneous beating of cardiac organoids, which typically becomes evident under a light microscope.
   4. By Day 12+, confirm that contractile rhythm is regular and sustained.
      NOTE: Beat rate and morphology can be quantified using timelapse microscopy and image analysis tools.
      Electrophysiological analysis (e.g., calcium imaging or MEA) can be performed from Day 12 onwards to assess functional maturity.

3. Preparation for cryopreservation

1. Carefully transfer cardiac organoids from the culture plate using a wide-bore pipette to avoid mechanical stress. Place them in a 15 mL conical tube and allow the organoids to settle naturally by gravity for 3-5 min without centrifugation.
2. Prepare Cryopreservation Medium (Cryo-CM) with the following components (final concentrations): 80% CM-M, 10% FBS, 10% DMSO.
3. Gently resuspend cardiac organoids in 500 µL of Cryo-CM per cryovial.
4. Place cryovials into an isopropanol-based freezing container. Store at -80 °C for 12-24 h to ensure gradual cooling at ~1 °C/min.
5. After overnight freezing, transfer vials into liquid nitrogen storage for long-term preservation.
   NOTE: We cryopreserved the cardiac organoids at a density of approximately 20-30 mature organoids per cryovial, suspended in 1 mL of cryopreservation medium.

4. Cardiac organoid recovery and thawing

1. To recover organoids, quickly thaw vials in a water bath, transfer the cardiac organoids to CM-M medium, gently centrifuge to remove DMSO, and resuspend the organoids in fresh CM-M..
   NOTE: Postthaw beating typically resumes within 24-48 h, depending on the maturity of organoids and handling quality.

The successful differentiation of H9 hESCs into cardiac organoids begins with the robust revival and maintenance of pluripotent stem cells. Upon thawing, hESCs demonstrate high viability and adherence when plated on vitronectin-coated plates in E8 medium supplemented with a ROCK inhibitor. Within 24-48 h, compact colonies with defined edges and high nucleus-to-cytoplasm ratios can be observed, indicating recovery of typical undifferentiated morphology (Figure 1A). After thawing, hESCs rapidly enter a proliferative phase, demonstrating robust colony expansion and re-establishment of pluripotent characteristics (Figure 1A).

The overall workflow for generating cardiac organoids from H9 hESCs is illustrated in Figure 1B, detailing each stage from stem cell maintenance to 3D organoid formation and maturation. The successful differentiation of hESCs into cardiac organoids is evident through a series of distinct morphological and functional transitions. Within the first 24 h following aggregation in ultra-low attachment plates, compact and uniform spheroids form. These spheroids exhibit a smooth, round morphology with high optical density, indicating high cell viability and homogeneous aggregation (Figure 1C).

By Day 4-5 of mesodermal induction, the spheroids begin to display a flattened and slightly expanded appearance, reflecting cellular differentiation and structural remodeling (Figure 1C). From Day 7 onward, spontaneous contractions are typically observed in over 70% of aggregates, indicating the onset of cardiomyocyte functionality (Figure 1C and Video 1).

By Day 12+, mature cardiac organoids demonstrate synchronous and rhythmic contractions. Immunofluorescence staining confirms the expression of the cardiac-specific marker cardiac troponin T (cTnT), which is distributed throughout the cardiac organoids (Figure 2A), indicating the presence of functionally relevant cardiomyocyte populations. Three-dimensional immunofluorescence staining further demonstrates the presence of vascular and smooth muscle lineage components within the cardiac organoids. CDH5 (VE-cadherin), a marker of endothelial cells, is broadly detected across the organoid surface and within interior regions, suggesting endothelial network formation. Concurrently, α-smooth muscle actin (αSMA) expression is observed in a subset of cells (Figure 2B). The co-expression of αSMA and CDH5 in spatially distinct domains highlights the multicellular complexity and tissue-like architecture of the cardiac organoids. In addition, quantitative analysis reveals that approximately 63% of cells within the cardiac organoids are positive for cTnT, 17% are positive for αSMA, and 9% express CDH5. These proportions are consistently observed across three independent experimental batches, demonstrating the robustness and reproducibility of the differentiation protocol.

Following isoproterenol (ISO)-induced injury, cardiac organoids exhibited altered expression of several cardiac markers. Quantitative PCR analysis showed a notably higher expression of MYH7 compared to MYH6 (Figure 3), consistent with a ventricular-like transcriptional profile. Additionally, elevated levels of ANP and BNP mRNA suggested activation of cardiac stress pathways. These results suggest that the organoids respond physiologically to injury and contain functionally relevant cardiomyocyte subtypes.

Organoids subjected to drug treatments or electrical stimulation may display altered beating rates or arrhythmic patterns, providing a platform for pharmacological testing or disease modeling. Moreover, cryopreserved organoids retain beating capacity upon thawing, though a brief recovery period (24-48 h) is required before functional assessments can be made.

Figure 1: Establishment and morphological progression of cardiac organoids from hESCs (H9). (A) Representative phase-contrast images showing the recovery process of hESCs at day 1, day 2, and day 3 post-thawing.(B) Schematic workflow outlining the stepwise differentiation protocol used for the generation of cardiac organoids from hESCs.(C) Brightfield images showing the morphological characteristics of cardiac organoids at key differentiation stages. (D) Spontaneous beating frequency of cardiac organoids. Cardiac organoids derived from H9 hESCs were monitored under brightfield time-lapse microscopy. Beating frequency was quantified on Day 12 by counting rhythmic contractions per minute in individual organoids. Data are presented as beats per minute (bpm), with each dot representing a single organoid. Error bars indicate mean ± standard error of the mean (SEM). Scale bars = 100 µm (A,C). Abbreviation: hESCs = human embryonic stem cells. Please click here to view a larger version of this figure.

Figure 2: Immunofluorescence characterization of cardiac organoids. (A) Immunofluorescence staining for cTnT demonstrates the presence of cardiomyocytes within the cardiac organoids at day 12.(B) 3D immunofluorescence staining shows positive expression of aSMA and CDH5, indicating the emergence of smooth muscle and endothelial-like cell populations, respectively. Scale bars = 100 µm (A), 50 µm (B). Abbreviations: cTnT = cardiac troponin T; aSMA = smooth muscle actin alpha; CDH5 = cadherin-5. Please click here to view a larger version of this figure.

Figure 3. Expression of cardiac marker genes MYH6, MYH7, ANP, and BNP in cardiac organoids following ISO-induced injury. Quantitative PCR data are shown as mean ± SEM (n = 3 biological replicates). Statistical analysis was performed using one-way ANOVA followed by Dunnett's post hoc test for multiple comparisons. P < 0.05 was considered statistically significant (*P < 0.05, **P < 0.01, ***P < 0.001). Abbreviation: ISO = isoproterenol. Please click here to view a larger version of this figure.

Video 1: Spontaneous contractile activity of cardiac organoids. Please click here to download this File.

The successful generation of cardiac organoids from H9 hESCs relies on precise temporal and spatial regulation of key signaling pathways to mimic early cardiac development. This protocol recapitulates the critical stages of mesoderm induction, cardiac lineage specification, and self-organization into 3D contractile structures in a chemically defined system. The method provides a robust and reproducible platform for modeling human heart development, congenital heart disease, and cardiotoxicity screening.

One of the most critical steps in this protocol is ensuring the health and quality of the H9 hESCs prior to differentiation 5. Cells must be cultured to approximately 70% confluency and passaged at least once after thawing before differentiation is initiated^15,16. Using freshly thawed hESCs without passaging often results in poor viability and inefficient cardiac induction. Additionally, during routine maintenance, overdigestion or excessive pipetting should be avoided to preserve colony integrity; optimal results are achieved when cells are maintained in small clusters of 2-10 cells. Avoid enzymatic passaging to maintain pluripotency and colony morphology. Use the stem cell dissociation reagent to preserve cell-cell contact and minimize stress, but avoid overdigestion or excessive pipetting that can result in a single-cell suspension.

To ensure reproducibility and scalability, several experimental constraints and quality control measures must be considered. In particular, cell viability should exceed 85% post-thaw, and differentiation should not be attempted directly after thawing. Cells should undergo at least one passage before induction. Uniform spheroid formation is essential; compact embryoid bodies (EBs) should form within 24 h of aggregation in ultra-low attachment plates. Spontaneous beating typically begins by Day 7.5, and failure to observe contractions in >70% of organoids may indicate protocol deviation or suboptimal cell quality. These metrics help standardize outcomes across laboratories and facilitate broader adoption of this cardiac organoid model.

Another essential factor is the precise composition and timing of media components. MD-M, enriched with Activin A, BMP4, FGF2, CHIR99021, and a PI3K inhibitor, initiates efficient induction of mesodermal fate⁷. The subsequent stages use Wnt inhibitors and retinoic acid to promote cardiac specification and maturation. The use of a chemically defined medium throughout the protocol minimizes batch variability and supports consistent cardiac differentiation outcomes⁵.

Unlike other organoid systems, cardiac organoids do not support passaging or serial expansion^5,17, mainly due to the limited proliferative capacity of cardiomyocytes, even at immature stages. Attempts to passage or mechanically dissociate cardiac organoids often result in loss of contractile function and structural integrity. Therefore, it is recommended that aggregates be formed at optimized cell densities and maintained without physical disturbance during the beating stages.

Compared to existing cardiac organoid protocols⁹, this method offers several notable advantages. First, it utilizes a chemically defined culture system with standardized and commercially available components, thereby minimizing batch-to-batch and inter-experimental variability and enhancing reproducibility. Second, the protocol supports cryopreservation of cardiac organoids with robust post-thaw functional recovery, enabling long-term storage and longitudinal analysis. Collectively, these features contribute to the reliability and consistency of cardiac organoid generation, making this protocol particularly suitable for scalable applications and comparative studies.

In addition, it avoids the use of undefined factors such as serum or feeder layers, enabling more standardized and scalable cardiac differentiation. It faithfully generates 3D structures that spontaneously contract and express cardiac markers, providing a physiologically relevant system for studying human-specific cardiac development. The ability to cryopreserve and recover beating cardiac organoids further enhances the versatility of this system for longitudinal studies or high-throughput applications.

In this study, cardiac organoids were cryopreserved using a specialized freezing medium and a standard slow-freezing protocol. Upon thawing, the organoids were maintained in maturation medium for continued culture. Survival rate was assessed 48 h post thaw by calculating the proportion of organoids exhibiting both intact morphology and spontaneous beating, as observed under a light microscope. Although we did not directly compare the proportions of cardiomyocytes, smooth muscle cells, and endothelial cells before and after freezing, we observed that the vast majority of cardiac organoids gradually regained spontaneous contractile activity within 3-4 days post thaw. Furthermore, they maintained visible beating behavior throughout the approximately 14 day observation period. However, compared to the pre-freezing state, thawed organoids generally exhibited a reduction in both beating frequency and contractile strength, indicating a trend toward diminished functional activity. These findings suggest that, although the cryopreserved organoids retain basic structural integrity and functionality after recovery, their physiological activity may be partially compromised. This highlights the need for further optimization of cryopreservation strategies to better preserve functional performance.

Despite its robustness, this protocol is not without limitations. Some degree of variability may arise due to differences in cell line behavior or environmental factors such as CO₂ levels and humidity. In addition, long-term culture beyond 12 days may lead to central necrosis or fibrosis-like features due to limited nutrient and oxygen diffusion, especially in the absence of vascularization. Co-culture with endothelial cells or application of microfluidic perfusion systems may improve these limitations in future adaptations. Further, these cardiac organoids consistently exhibited rhythmic and spontaneous contractile activity, suggesting the presence of functional cardiomyocyte populations; however, direct electrophysiological assessments to confirm definitive electromechanical coupling were not conducted in the current study. Moreover, this study was primarily designed and optimized using the H9 hESC line, which is a widely accepted standard model in the field of cardiac differentiation and has been extensively validated for its reliable differentiation efficiency and stability. Therefore, at this stage of the research, we did not conduct validation experiments using additional cell lines. In future studies, we plan to extend this work to include other hESC or induced pluripotent stem cell (iPSC) lines to further enhance the generalizability and applicability of our findings.

While we did not perform systematic subtype characterization of atrial versus ventricular cardiomyocytes, our ISO-injury experiments revealed a significantly higher MYH7-to-MYH6 expression ratio, consistent with a ventricular-like profile, alongside increased ANP and BNP expression as markers of cardiac stress. These results indicate that the cardiac organoids not only contain functional cardiomyocyte populations but also exhibit physiologically relevant stress responses. Nonetheless, more comprehensive analyses, including immunostaining and electrophysiological profiling, will be required to fully delineate subtype identities and functional heterogeneity in future studies.

Although cryopreserved organoids demonstrated survival and spontaneous beating upon recovery, a systematic comparison of cellular composition before and after freezing was not performed in this study. This represents a limitation, as potential shifts in the relative proportions of cardiomyocytes, smooth muscle cells, and endothelial-like populations may have occurred but were not quantified. Future studies incorporating flow cytometry, immunostaining, or single-cell RNA sequencing will be necessary to rigorously evaluate the impact of cryopreservation on cell-type composition and functional integrity.

In conclusion, this method provides a simple, reproducible, and efficient strategy for generating human cardiac organoids from pluripotent stem cells. It is a valuable tool for developmental biology, disease modeling, and cardiotoxicity studies. Future improvements, such as the incorporation of vascular networks or innervation, may enhance the physiological relevance of the model even further.